Merge git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable

* git://git.kernel.org/pub/scm/linux/kernel/git/mason/btrfs-unstable:
  Btrfs: fix fiemap bugs with delalloc
  Btrfs: set FMODE_EXCL in btrfs_device->mode
  Btrfs: make btrfs_rm_device() fail gracefully
  Btrfs: Avoid accessing unmapped kernel address
  Btrfs: Fix BTRFS_IOC_SUBVOL_SETFLAGS ioctl
  Btrfs: allow balance to explicitly allocate chunks as it relocates
  Btrfs: put ENOSPC debugging under a mount option

fs/btrfs/ctree.h

@@ -1254,6 +1254,7 @@ struct btrfs_root {
 #define BTRFS_MOUNT_SPACE_CACHE		(1 << 12)
 #define BTRFS_MOUNT_CLEAR_CACHE		(1 << 13)
 #define BTRFS_MOUNT_USER_SUBVOL_RM_ALLOWED (1 << 14)
+#define BTRFS_MOUNT_ENOSPC_DEBUG	 (1 << 15)
 
 #define btrfs_clear_opt(o, opt)		((o) &= ~BTRFS_MOUNT_##opt)
 #define btrfs_set_opt(o, opt)		((o) |= BTRFS_MOUNT_##opt)
@@ -2218,6 +2219,8 @@ int btrfs_error_unpin_extent_range(struct btrfs_root *root,
 				   u64 start, u64 end);
 int btrfs_error_discard_extent(struct btrfs_root *root, u64 bytenr,
 			       u64 num_bytes);
+int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root, u64 type);
 
 /* ctree.c */
 int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key,

fs/btrfs/extent-tree.c

@@ -5376,7 +5376,7 @@ int btrfs_reserve_extent(struct btrfs_trans_handle *trans,
 			       num_bytes, data, 1);
 		goto again;
 	}
-	if (ret == -ENOSPC) {
+	if (ret == -ENOSPC && btrfs_test_opt(root, ENOSPC_DEBUG)) {
 		struct btrfs_space_info *sinfo;
 
 		sinfo = __find_space_info(root->fs_info, data);
@@ -8065,6 +8065,13 @@ int btrfs_set_block_group_ro(struct btrfs_root *root,
 	return ret;
 }
 
+int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans,
+			    struct btrfs_root *root, u64 type)
+{
+	u64 alloc_flags = get_alloc_profile(root, type);
+	return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1);
+}
+
 /*
  * helper to account the unused space of all the readonly block group in the
  * list. takes mirrors into account.

fs/btrfs/extent_io.c

@@ -1433,12 +1433,13 @@ int extent_clear_unlock_delalloc(struct inode *inode,
  */
 u64 count_range_bits(struct extent_io_tree *tree,
 		     u64 *start, u64 search_end, u64 max_bytes,
-		     unsigned long bits)
+		     unsigned long bits, int contig)
 {
 	struct rb_node *node;
 	struct extent_state *state;
 	u64 cur_start = *start;
 	u64 total_bytes = 0;
+	u64 last = 0;
 	int found = 0;
 
 	if (search_end <= cur_start) {
@@ -1463,7 +1464,9 @@ u64 count_range_bits(struct extent_io_tree *tree,
 		state = rb_entry(node, struct extent_state, rb_node);
 		if (state->start > search_end)
 			break;
-		if (state->end >= cur_start && (state->state & bits)) {
+		if (contig && found && state->start > last + 1)
+			break;
+		if (state->end >= cur_start && (state->state & bits) == bits) {
 			total_bytes += min(search_end, state->end) + 1 -
 				       max(cur_start, state->start);
 			if (total_bytes >= max_bytes)
@@ -1472,6 +1475,9 @@ u64 count_range_bits(struct extent_io_tree *tree,
 				*start = state->start;
 				found = 1;
 			}
+			last = state->end;
+		} else if (contig && found) {
+			break;
 		}
 		node = rb_next(node);
 		if (!node)
@@ -2912,6 +2918,46 @@ sector_t extent_bmap(struct address_space *mapping, sector_t iblock,
 	return sector;
 }
 
+/*
+ * helper function for fiemap, which doesn't want to see any holes.
+ * This maps until we find something past 'last'
+ */
+static struct extent_map *get_extent_skip_holes(struct inode *inode,
+						u64 offset,
+						u64 last,
+						get_extent_t *get_extent)
+{
+	u64 sectorsize = BTRFS_I(inode)->root->sectorsize;
+	struct extent_map *em;
+	u64 len;
+
+	if (offset >= last)
+		return NULL;
+
+	while(1) {
+		len = last - offset;
+		if (len == 0)
+			break;
+		len = (len + sectorsize - 1) & ~(sectorsize - 1);
+		em = get_extent(inode, NULL, 0, offset, len, 0);
+		if (!em || IS_ERR(em))
+			return em;
+
+		/* if this isn't a hole return it */
+		if (!test_bit(EXTENT_FLAG_VACANCY, &em->flags) &&
+		    em->block_start != EXTENT_MAP_HOLE) {
+			return em;
+		}
+
+		/* this is a hole, advance to the next extent */
+		offset = extent_map_end(em);
+		free_extent_map(em);
+		if (offset >= last)
+			break;
+	}
+	return NULL;
+}
+
 int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len, get_extent_t *get_extent)
 {
@@ -2921,16 +2967,19 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	u32 flags = 0;
 	u32 found_type;
 	u64 last;
+	u64 last_for_get_extent = 0;
 	u64 disko = 0;
+	u64 isize = i_size_read(inode);
 	struct btrfs_key found_key;
 	struct extent_map *em = NULL;
 	struct extent_state *cached_state = NULL;
 	struct btrfs_path *path;
 	struct btrfs_file_extent_item *item;
 	int end = 0;
-	u64 em_start = 0, em_len = 0;
+	u64 em_start = 0;
+	u64 em_len = 0;
+	u64 em_end = 0;
 	unsigned long emflags;
-	int hole = 0;
 
 	if (len == 0)
 		return -EINVAL;
@@ -2940,6 +2989,10 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		return -ENOMEM;
 	path->leave_spinning = 1;
 
+	/*
+	 * lookup the last file extent.  We're not using i_size here
+	 * because there might be preallocation past i_size
+	 */
 	ret = btrfs_lookup_file_extent(NULL, BTRFS_I(inode)->root,
 				       path, inode->i_ino, -1, 0);
 	if (ret < 0) {
@@ -2953,18 +3006,38 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
 	found_type = btrfs_key_type(&found_key);
 
-	/* No extents, just return */
+	/* No extents, but there might be delalloc bits */
 	if (found_key.objectid != inode->i_ino ||
 	    found_type != BTRFS_EXTENT_DATA_KEY) {
-		btrfs_free_path(path);
-		return 0;
+		/* have to trust i_size as the end */
+		last = (u64)-1;
+		last_for_get_extent = isize;
+	} else {
+		/*
+		 * remember the start of the last extent.  There are a
+		 * bunch of different factors that go into the length of the
+		 * extent, so its much less complex to remember where it started
+		 */
+		last = found_key.offset;
+		last_for_get_extent = last + 1;
 	}
-	last = found_key.offset;
 	btrfs_free_path(path);
 
+	/*
+	 * we might have some extents allocated but more delalloc past those
+	 * extents.  so, we trust isize unless the start of the last extent is
+	 * beyond isize
+	 */
+	if (last < isize) {
+		last = (u64)-1;
+		last_for_get_extent = isize;
+	}
+
 	lock_extent_bits(&BTRFS_I(inode)->io_tree, start, start + len, 0,
 			 &cached_state, GFP_NOFS);
-	em = get_extent(inode, NULL, 0, off, max - off, 0);
+
+	em = get_extent_skip_holes(inode, off, last_for_get_extent,
+				   get_extent);
 	if (!em)
 		goto out;
 	if (IS_ERR(em)) {
@@ -2973,19 +3046,14 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 	}
 
 	while (!end) {
-		hole = 0;
-		off = em->start + em->len;
+		off = extent_map_end(em);
 		if (off >= max)
 			end = 1;
 
-		if (em->block_start == EXTENT_MAP_HOLE) {
-			hole = 1;
-			goto next;
-		}
-
 		em_start = em->start;
 		em_len = em->len;
-
+		em_end = extent_map_end(em);
+		emflags = em->flags;
 		disko = 0;
 		flags = 0;
 
@@ -3004,37 +3072,29 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags))
 			flags |= FIEMAP_EXTENT_ENCODED;
 
-next:
-		emflags = em->flags;
 		free_extent_map(em);
 		em = NULL;
-		if (!end) {
-			em = get_extent(inode, NULL, 0, off, max - off, 0);
-			if (!em)
-				goto out;
-			if (IS_ERR(em)) {
-				ret = PTR_ERR(em);
-				goto out;
-			}
-			emflags = em->flags;
-		}
-
-		if (test_bit(EXTENT_FLAG_VACANCY, &emflags)) {
+		if ((em_start >= last) || em_len == (u64)-1 ||
+		   (last == (u64)-1 && isize <= em_end)) {
 			flags |= FIEMAP_EXTENT_LAST;
 			end = 1;
 		}
 
-		if (em_start == last) {
+		/* now scan forward to see if this is really the last extent. */
+		em = get_extent_skip_holes(inode, off, last_for_get_extent,
+					   get_extent);
+		if (IS_ERR(em)) {
+			ret = PTR_ERR(em);
+			goto out;
+		}
+		if (!em) {
 			flags |= FIEMAP_EXTENT_LAST;
 			end = 1;
 		}
-
-		if (!hole) {
-			ret = fiemap_fill_next_extent(fieinfo, em_start, disko,
-						em_len, flags);
-			if (ret)
-				goto out_free;
-		}
+		ret = fiemap_fill_next_extent(fieinfo, em_start, disko,
+					      em_len, flags);
+		if (ret)
+			goto out_free;
 	}
 out_free:
 	free_extent_map(em);

fs/btrfs/extent_io.h

@@ -191,7 +191,7 @@ void extent_io_exit(void);
 
 u64 count_range_bits(struct extent_io_tree *tree,
 		     u64 *start, u64 search_end,
-		     u64 max_bytes, unsigned long bits);
+		     u64 max_bytes, unsigned long bits, int contig);
 
 void free_extent_state(struct extent_state *state);
 int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end,

fs/btrfs/inode.c

@@ -1913,7 +1913,7 @@ static int btrfs_clean_io_failures(struct inode *inode, u64 start)
 
 	private = 0;
 	if (count_range_bits(&BTRFS_I(inode)->io_failure_tree, &private,
-			     (u64)-1, 1, EXTENT_DIRTY)) {
+			     (u64)-1, 1, EXTENT_DIRTY, 0)) {
 		ret = get_state_private(&BTRFS_I(inode)->io_failure_tree,
 					start, &private_failure);
 		if (ret == 0) {
@@ -5280,6 +5280,128 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page,
 	return em;
 }
 
+struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *page,
+					   size_t pg_offset, u64 start, u64 len,
+					   int create)
+{
+	struct extent_map *em;
+	struct extent_map *hole_em = NULL;
+	u64 range_start = start;
+	u64 end;
+	u64 found;
+	u64 found_end;
+	int err = 0;
+
+	em = btrfs_get_extent(inode, page, pg_offset, start, len, create);
+	if (IS_ERR(em))
+		return em;
+	if (em) {
+		/*
+		 * if our em maps to a hole, there might
+		 * actually be delalloc bytes behind it
+		 */
+		if (em->block_start != EXTENT_MAP_HOLE)
+			return em;
+		else
+			hole_em = em;
+	}
+
+	/* check to see if we've wrapped (len == -1 or similar) */
+	end = start + len;
+	if (end < start)
+		end = (u64)-1;
+	else
+		end -= 1;
+
+	em = NULL;
+
+	/* ok, we didn't find anything, lets look for delalloc */
+	found = count_range_bits(&BTRFS_I(inode)->io_tree, &range_start,
+				 end, len, EXTENT_DELALLOC, 1);
+	found_end = range_start + found;
+	if (found_end < range_start)
+		found_end = (u64)-1;
+
+	/*
+	 * we didn't find anything useful, return
+	 * the original results from get_extent()
+	 */
+	if (range_start > end || found_end <= start) {
+		em = hole_em;
+		hole_em = NULL;
+		goto out;
+	}
+
+	/* adjust the range_start to make sure it doesn't
+	 * go backwards from the start they passed in
+	 */
+	range_start = max(start,range_start);
+	found = found_end - range_start;
+
+	if (found > 0) {
+		u64 hole_start = start;
+		u64 hole_len = len;
+
+		em = alloc_extent_map(GFP_NOFS);
+		if (!em) {
+			err = -ENOMEM;
+			goto out;
+		}
+		/*
+		 * when btrfs_get_extent can't find anything it
+		 * returns one huge hole
+		 *
+		 * make sure what it found really fits our range, and
+		 * adjust to make sure it is based on the start from
+		 * the caller
+		 */
+		if (hole_em) {
+			u64 calc_end = extent_map_end(hole_em);
+
+			if (calc_end <= start || (hole_em->start > end)) {
+				free_extent_map(hole_em);
+				hole_em = NULL;
+			} else {
+				hole_start = max(hole_em->start, start);
+				hole_len = calc_end - hole_start;
+			}
+		}
+		em->bdev = NULL;
+		if (hole_em && range_start > hole_start) {
+			/* our hole starts before our delalloc, so we
+			 * have to return just the parts of the hole
+			 * that go until  the delalloc starts
+			 */
+			em->len = min(hole_len,
+				      range_start - hole_start);
+			em->start = hole_start;
+			em->orig_start = hole_start;
+			/*
+			 * don't adjust block start at all,
+			 * it is fixed at EXTENT_MAP_HOLE
+			 */
+			em->block_start = hole_em->block_start;
+			em->block_len = hole_len;
+		} else {
+			em->start = range_start;
+			em->len = found;
+			em->orig_start = range_start;
+			em->block_start = EXTENT_MAP_DELALLOC;
+			em->block_len = found;
+		}
+	} else if (hole_em) {
+		return hole_em;
+	}
+out:
+
+	free_extent_map(hole_em);
+	if (err) {
+		free_extent_map(em);
+		return ERR_PTR(err);
+	}
+	return em;
+}
+
 static struct extent_map *btrfs_new_extent_direct(struct inode *inode,
 						  u64 start, u64 len)
 {
@@ -6102,7 +6224,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
 static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
 		__u64 start, __u64 len)
 {
-	return extent_fiemap(inode, fieinfo, start, len, btrfs_get_extent);
+	return extent_fiemap(inode, fieinfo, start, len, btrfs_get_extent_fiemap);
 }
 
 int btrfs_readpage(struct file *file, struct page *page)

fs/btrfs/ioctl.c

@@ -1071,12 +1071,15 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
 	if (copy_from_user(&flags, arg, sizeof(flags)))
 		return -EFAULT;
 
-	if (flags & ~BTRFS_SUBVOL_CREATE_ASYNC)
+	if (flags & BTRFS_SUBVOL_CREATE_ASYNC)
 		return -EINVAL;
 
 	if (flags & ~BTRFS_SUBVOL_RDONLY)
 		return -EOPNOTSUPP;
 
+	if (!is_owner_or_cap(inode))
+		return -EACCES;
+
 	down_write(&root->fs_info->subvol_sem);
 
 	/* nothing to do */
@@ -1097,7 +1100,7 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file,
 		goto out_reset;
 	}
 
-	ret = btrfs_update_root(trans, root,
+	ret = btrfs_update_root(trans, root->fs_info->tree_root,
 				&root->root_key, &root->root_item);
 
 	btrfs_commit_transaction(trans, root);

fs/btrfs/lzo.c

@@ -280,6 +280,7 @@ static int lzo_decompress_biovec(struct list_head *ws,
 	unsigned long tot_out;
 	unsigned long tot_len;
 	char *buf;
+	bool may_late_unmap, need_unmap;
 
 	data_in = kmap(pages_in[0]);
 	tot_len = read_compress_length(data_in);
@@ -300,11 +301,13 @@ static int lzo_decompress_biovec(struct list_head *ws,
 
 		tot_in += in_len;
 		working_bytes = in_len;
+		may_late_unmap = need_unmap = false;
 
 		/* fast path: avoid using the working buffer */
 		if (in_page_bytes_left >= in_len) {
 			buf = data_in + in_offset;
 			bytes = in_len;
+			may_late_unmap = true;
 			goto cont;
 		}
 
@@ -329,14 +332,17 @@ static int lzo_decompress_biovec(struct list_head *ws,
 				if (working_bytes == 0 && tot_in >= tot_len)
 					break;
 
-				kunmap(pages_in[page_in_index]);
-				page_in_index++;
-				if (page_in_index >= total_pages_in) {
+				if (page_in_index + 1 >= total_pages_in) {
 					ret = -1;
-					data_in = NULL;
 					goto done;
 				}
-				data_in = kmap(pages_in[page_in_index]);
+
+				if (may_late_unmap)
+					need_unmap = true;
+				else
+					kunmap(pages_in[page_in_index]);
+
+				data_in = kmap(pages_in[++page_in_index]);
 
 				in_page_bytes_left = PAGE_CACHE_SIZE;
 				in_offset = 0;
@@ -346,6 +352,8 @@ static int lzo_decompress_biovec(struct list_head *ws,
 		out_len = lzo1x_worst_compress(PAGE_CACHE_SIZE);
 		ret = lzo1x_decompress_safe(buf, in_len, workspace->buf,
 					    &out_len);
+		if (need_unmap)
+			kunmap(pages_in[page_in_index - 1]);
 		if (ret != LZO_E_OK) {
 			printk(KERN_WARNING "btrfs decompress failed\n");
 			ret = -1;
@@ -363,8 +371,7 @@ static int lzo_decompress_biovec(struct list_head *ws,
 			break;
 	}
 done:
-	if (data_in)
-		kunmap(pages_in[page_in_index]);
+	kunmap(pages_in[page_in_index]);
 	return ret;
 }
 

fs/btrfs/relocation.c

@@ -3654,6 +3654,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 	u32 item_size;
 	int ret;
 	int err = 0;
+	int progress = 0;
 
 	path = btrfs_alloc_path();
 	if (!path)
@@ -3666,9 +3667,10 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 	}
 
 	while (1) {
+		progress++;
 		trans = btrfs_start_transaction(rc->extent_root, 0);
 		BUG_ON(IS_ERR(trans));
-
+restart:
 		if (update_backref_cache(trans, &rc->backref_cache)) {
 			btrfs_end_transaction(trans, rc->extent_root);
 			continue;
@@ -3781,6 +3783,15 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc)
 			}
 		}
 	}
+	if (trans && progress && err == -ENOSPC) {
+		ret = btrfs_force_chunk_alloc(trans, rc->extent_root,
+					      rc->block_group->flags);
+		if (ret == 0) {
+			err = 0;
+			progress = 0;
+			goto restart;
+		}
+	}
 
 	btrfs_release_path(rc->extent_root, path);
 	clear_extent_bits(&rc->processed_blocks, 0, (u64)-1, EXTENT_DIRTY,

fs/btrfs/super.c

@@ -155,7 +155,8 @@ enum {
 	Opt_nossd, Opt_ssd_spread, Opt_thread_pool, Opt_noacl, Opt_compress,
 	Opt_compress_type, Opt_compress_force, Opt_compress_force_type,
 	Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard,
-	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, Opt_err,
+	Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed,
+	Opt_enospc_debug, Opt_err,
 };
 
 static match_table_t tokens = {
@@ -184,6 +185,7 @@ static match_table_t tokens = {
 	{Opt_space_cache, "space_cache"},
 	{Opt_clear_cache, "clear_cache"},
 	{Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"},
+	{Opt_enospc_debug, "enospc_debug"},
 	{Opt_err, NULL},
 };
 
@@ -358,6 +360,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options)
 		case Opt_user_subvol_rm_allowed:
 			btrfs_set_opt(info->mount_opt, USER_SUBVOL_RM_ALLOWED);
 			break;
+		case Opt_enospc_debug:
+			btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG);
+			break;
 		case Opt_err:
 			printk(KERN_INFO "btrfs: unrecognized mount option "
 			       "'%s'\n", p);

fs/btrfs/volumes.c

@@ -1338,11 +1338,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 
 	ret = btrfs_shrink_device(device, 0);
 	if (ret)
-		goto error_brelse;
+		goto error_undo;
 
 	ret = btrfs_rm_dev_item(root->fs_info->chunk_root, device);
 	if (ret)
-		goto error_brelse;
+		goto error_undo;
 
 	device->in_fs_metadata = 0;
 
@@ -1416,6 +1416,13 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path)
 	mutex_unlock(&root->fs_info->volume_mutex);
 	mutex_unlock(&uuid_mutex);
 	return ret;
+error_undo:
+	if (device->writeable) {
+		list_add(&device->dev_alloc_list,
+			 &root->fs_info->fs_devices->alloc_list);
+		root->fs_info->fs_devices->rw_devices++;
+	}
+	goto error_brelse;
 }
 
 /*
@@ -1633,7 +1640,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path)
 	device->dev_root = root->fs_info->dev_root;
 	device->bdev = bdev;
 	device->in_fs_metadata = 1;
-	device->mode = 0;
+	device->mode = FMODE_EXCL;
 	set_blocksize(device->bdev, 4096);
 
 	if (seeding_dev) {